Security Week 03: SHA-1 sunset continues, bug or feature in Whatsapp, vulnerabilities in routers not fixed

    SHA-1 is everything. Or not? Keeping track of developments around this hashing algorithm is easy and pleasant: despite the obvious seriousness of the problem, it remains of little use for practical attacks, at least mass ones. The first time I mentioned SHA-1 in a digest from October 2015. Problems with it appeared due to the fact that computing resources became cheaper somewhat faster than expected. Cryptography expert Bruce Schneier predicted in 2012 that in three years it would take 11 years of computing on a conditional server to create a collision when generating hashes. Three years passed, and it turned out that in fact (due to the development of parallel computing technologies, and thanks to new research in the field of cryptography) this period is much shorter - only 49 days.

    Since SHA-1 hashing is used in highly critical operations, for example, when establishing a secure connection to websites, software developers quite quickly began to share plans for decommissioning an unreliable algorithm. Starting January 24 (for Firefox, for other browsers a bit later), visiting a site that does not support the more robust SHA-2 algorithm (usually modifications to SHA-256) will lead to a variety of threatening warnings for users.

    According to Facebook at the end of 2015, 7% of their audience, mostly residents of developing countries, still used browsers or applications that did not support SHA-2. More recent research showsthat unreliable SHA-1s use 35% of sites in the IPv4 address space, but almost none of them are really popular. The most serious problems did not arise with sites and users, but with payment systems and mobile application developers. The main conclusion of a fairly detailed review on Threatpost: despite the early detection of the problem, public discussion of solutions, joint efforts, etc., in 2017 SHA-1 will become a niche problem that is unlikely to be finally solved. So, there will always be a risk of malicious manipulations with a half-forgotten server, which they either forgot to update or failed. You cannot just take and root out an unreliable algorithm.

    Whatsapp found a message encryption system vulnerability. Or not found.
    The news . Blog post with details.

    On Friday the 13th, an article was published in the Guardian newspaper that the vulnerability in Whatsapp allows messenger developers to eavesdrop on users with encryption of correspondence. The vulnerability details are as follows: The Signal protocol, on which Whatsapp encryption is based, generates unique keys for communication between two users. In the event that the user is offline, Whatsapp can change the keys, and quietly, without notifying users - if you do not enable the corresponding function in the settings.

    That is, if you change the keys, then for some reason you need it, but why else do you need it, except for eavesdropping? The developer of Open Whisper Systems, responsible for the Signal protocol, Morley Marlinspike tried to destroy this ironic media logic. The situation of replacing the encryption key in question occurs if the server has nowhere to deliver the message. For example, a smartphone has been replaced or the application has been reinstalled. In this case, a new key is generated on the server to ensure message delivery when the user returns to the network.

    In short, the bug turned out to be a feature. The question is why users are not notified about the change of encryption keys? Whatsapp developers felt that users, most of whom do not understand anything in encryption, should not be scared again. In the Signal messenger, for example, they consider that it is worth scaring, and in Whatsapp this setting can be turned on manually. Then you will have the opportunity to turn on the paranoid mode: if (it doesn’t matter for some reason) you see that the keys have changed, you can change them again, presumably ensuring yourself the maximum privacy mode.

    Although it will not help the paranoid. Interesting news about the need for a neat interpretation of cybersecurity aspects.

    Routers of the Thai Internet provider remain vulnerable six months after the discovery of the vulnerability
    News . Research .

    Traditional news from the world of "fake IoT." The Thai Internet provider TrueOnline discovered vulnerable routers produced by the little-known company Billion and the famous Zyxel, in which researchers found a lot of ways to unauthorized interception of control via a web interface (including outside). Apparently, we are talking about vulnerabilities in custom firmware - vendors have nothing to do with it, although options are possible (Zyxel said that we are talking about routers that have long been discontinued)

    Vulnerabilities were found back in July this year, but they were published only now for a simple reason: the Internet provider completely ignored the reports of researchers. In general, routine news looks good in context. At the end of last year, Laboratory experts discovered ( news and research ) the Android Trojan Switcher, which has the function of replacing DNS records in routers, that is, it is used as a step to capture the entire victim’s home network.



    A resident very dangerous virus that infects .COM-, .EXE- and .SYS files that it searches in the current directory every time it invokes int 21h. .COM and .EXE files infect standardly. When an .SYS file is infected, it appends its body to the end of the file and modifies the Interrupt and Strategy programs of the infected driver.

    It is activated only in the absence of anti-virus blockers. When you try to pass the virus codes, the debugger destroys some of the data on the disk. Contains texts: "Dear Nina, you make me write this virus; Happy new year! ”,": \ COMMAND.COM ".

    Quote from the book "Computer viruses in MS-DOS" by Eugene Kaspersky. 1992 year. Page 41.

    Disclaimer:This column reflects only the private opinion of its author. It may coincide with the position of Kaspersky Lab, or it may not coincide. That's how lucky.

    Also popular now: